{
\def\sym#1{\ifmmode^{#1}\else\(^{#1}\)\fi}
\begin{tabular}{l*{9}{c}}
\hline\hline
                &Business Days         &Replied in time         &Complete answer         &Open request         &Extension was granted         &Replied within 15 bd         &Replied within 25 bd         &Replied in 15 bd         &Replied in 25 bd         \\
\hline
\vspace{-6mm}   &                  &                  &                  &                  &                  &                  &                  &                  &                  \\
[1em]
Treatment       &    1.992         &   -0.190\sym{**} &   -0.026         &   -0.005         &   -0.251\sym{***}&   -0.004         &   -0.080         &    0.007         &    0.075\sym{*}  \\
                &  (1.646)         &  (0.076)         &  (0.047)         &  (0.041)         &  (0.093)         &  (0.075)         &  (0.056)         &  (0.013)         &  (0.042)         \\
[1em]
Constant        &   20.159\sym{***}&    1.037\sym{***}&    1.055\sym{***}&   -0.028         &    1.050\sym{***}&    0.389\sym{***}&    1.025\sym{***}&    0.048         &    0.071         \\
                &  (2.435)         &  (0.106)         &  (0.079)         &  (0.062)         &  (0.126)         &  (0.096)         &  (0.088)         &  (0.031)         &  (0.057)         \\
\hline
\addlinespace[3pt]Observations&    2,077         &    2,164         &    2,164         &    2,164         &    2,164         &    2,164         &    2,164         &    2,164         &    2,164         \\
\addlinespace[3pt]R-squared&    0.277         &    0.243         &    0.241         &    0.424         &    0.354         &    0.298         &    0.188         &    0.025         &    0.054         \\
\hline\hline
\multicolumn{10}{l}{\footnotesize  {\begin{minipage}{1.4\textwidth} \small   \vspace{0.2cm}  \textit{Notes:} * p $ < $ 0.10, ** p $ < $ 0.05, *** p $ < $ 0.01. Standard errors clustered by agency are shown in parentheses. The dependent variable in the first column is the winsorized number of business days agencies took to respond (winsorized the 5\% highest values) and it only considers closed requests by November 27, 2019. Control variables include month in which the agency received the request, the total number of requests received by the agency before the intervention (Jan 1, 2017 until Sept. 5, 2018), the average percentage of pending requests (not past the deadline), and the average percentage of late requests prior to the intervention, type of requester, serial requesters, topic of the request, total number of requests received by the agency on the date request $ i $ arrived, and ministry of the agency.    \end{minipage}  } }\\
\end{tabular}
}
